Rubber data fusion and classification method based on terahertz time-domain spectroscopy technology

The invention discloses a black rubber type identification method based on a terahertz time-domain spectroscopy technology and a data fusion strategy, and the method comprises the following steps: (1) employing 8 different types of black rubber with a thickness of 1 mm, including natural rubber, chloroprene rubber, fluororubber, silicone rubber, butyl rubber, ethylene propylene diene monomer, nitrile rubber and styrene butadiene rubber, 8 samples are prepared from each kind of rubber; (2) putting each sample into a terahertz time-domain spectrometer, measuring in a transmission mode, taking four points for each sample, repeatedly measuring at each point for five times so as to ensure the reliability and accuracy of a data set, and then collecting terahertz spectrum data of the corresponding sample; (3) after the abnormal samples are removed, carrying out feature extraction on the terahertz absorption spectrum and refractive index spectrum data by adopting a wave band screening method, and meanwhile, extractin